Tech – Page 3

Google unveils Gemini 3 AI model and AI-first IDE called Antigravity

AI, AI coding, Artificial Intelligence, generative ai, Google, Google Gemini, Tech / Kris Guyer / November 18, 2025

Google’s flagship AI model is getting its second major upgrade this year.

Google has kicked its Gemini rollout into high gear over the past year, releasing the much-improved Gemini 2.5 family and cramming various flavors of the model into Search, Gmail, and just about everything else the company makes.

Now, Google’s increasingly unavoidable AI is getting an upgrade. Gemini 3 Pro is available in a limited form today, featuring more immersive, visual outputs and fewer lies, Google says. The company also says Gemini 3 sets a new high-water mark for vibe coding, and Google is announcing a new AI-first integrated development environment (IDE) called Antigravity, which is also available today.

The first member of the Gemini 3 family

Google says the release of Gemini 3 is yet another step toward artificial general intelligence (AGI). The new version of Google’s flagship AI model has expanded simulated reasoning abilities and shows improved understanding of text, images, and video. So far, testers like it—Google’s latest LLM is once again atop the LMArena leaderboard with an ELO score of 1,501, besting Gemini 2.5 Pro by 50 points.

Factuality has been a problem for all gen AI models, but Google says Gemini 3 is a big step in the right direction, and there are myriad benchmarks to tell the story. In the 1,000-question SimpleQA Verified test, Gemini 3 scored a record 72.1 percent. Yes, that means the state-of-the-art LLM still screws up almost 30 percent of general knowledge questions, but Google says this still shows substantial progress. On the much more difficult Humanity’s Last Exam, which tests PhD-level knowledge and reasoning, Gemini set another record, scoring 37.5 percent without tool use.

Math and coding are also a focus of Gemini 3. The model set new records in MathArena Apex (23.4 percent) and WebDev Arena (1487 ELO). In the SWE-bench Verified, which tests a model’s ability to generate code, Gemini 3 hit an impressive 76.2 percent.

So there are plenty of respectable but modest benchmark improvements, but Gemini 3 also won’t make you cringe as much. Google says it has tamped down on sycophancy, a common problem in all these overly polite LLMs. Outputs from Gemini 3 Pro are reportedly more concise, with less of what you want to hear and more of what you need to hear.

You can also expect Gemini 3 Pro to produce noticeably richer outputs. Google claims Gemini’s expanded reasoning capabilities keep it on task more effectively, allowing it to take action on your behalf. For example, Gemini 3 can triage and take action on your emails, creating to-do lists, summaries, recommended replies, and handy buttons to trigger suggested actions. This differs from the current Gemini models, which would only create a text-based to-do list with similar prompts.

The model also has what Google calls a “generative interface,” which comes in the form of two experimental output modes called visual layout and dynamic view. The former is a magazine-style interface that includes lots of images in a scrollable UI. Dynamic view leverages Gemini’s coding abilities to create custom interfaces—for example, a web app that explores the life and work of Vincent Van Gogh.

There will also be a Deep Think mode for Gemini 3, but that’s not ready for prime time yet. Google says it’s being tested by a small group for later release, but you should expect big things. Deep Think mode manages 41 percent in Humanity’s Last Exam without tools. Believe it or not, that’s an impressive score.

Coding with vibes

Google has offered several ways of generating and modifying code with Gemini models, but the launch of Gemini 3 adds a new one: Google Antigravity. This is Google’s new agentic development platform—it’s essentially an IDE designed around agentic AI, and it’s available in preview today.

With Antigravity, Google promises that you (the human) can get more work done by letting intelligent agents do the legwork. Google says you should think of Antigravity as a “mission control” for creating and monitoring multiple development agents. The AI in Antigravity can operate autonomously across the editor, terminal, and browser to create and modify projects, but everything they do is relayed to the user in the form of “Artifacts.” These sub-tasks are designed to be easily verifiable so you can keep on top of what the agent is doing. Gemini will be at the core of the Antigravity experience, but it’s not just Google’s bot. Antigravity also supports Claude Sonnet 4.5 and GPT-OSS agents.

Of course, developers can still plug into the Gemini API for coding tasks. With Gemini 3, Google is adding a client-side bash tool, which lets the AI generate shell commands in its workflow. The model can access file systems and automate operations, and a server-side bash tool will help generate code in multiple languages. This feature is starting in early access, though.

AI Studio is designed to be a faster way to build something with Gemini 3. Google says Gemini 3 Pro’s strong instruction following makes it the best vibe coding model yet, allowing non-programmers to create more complex projects.

A big experiment

Google will eventually have a whole family of Gemini 3 models, but there’s just the one for now. Gemini 3 Pro is rolling out in the Gemini app, AI Studio, Vertex AI, and the API starting today as an experiment. If you want to tinker with the new model in Google’s Antigravity IDE, that’s also available for testing today on Windows, Mac, and Linux.

Gemini 3 will also launch in the Google search experience on day one. You’ll have the option to enable Gemini 3 Pro in AI Mode, where Google says it will provide more useful information about a query. The generative interface capabilities from the Gemini app will be available here as well, allowing Gemini to create tools and simulations when appropriate to answer the user’s question. Google says these generative interfaces are strongly preferred in its user testing. This feature is available today, but only for AI Pro and Ultra subscribers.

Because the Pro model is the only Gemini 3 variant available in the preview, AI Overviews isn’t getting an immediate upgrade. That will come, but for now, Overviews will only reach out to Gemini 3 Pro for especially difficult search queries—basically the kind of thing Google thinks you should have used AI Mode to do in the first place.

There’s no official timeline for releasing more Gemini 3 models or graduating the Pro variant to general availability. However, given the wide rollout of the experimental release, it probably won’t be long.

Ryan Whitwam is a senior technology reporter at Ars Technica, covering the ways Google, AI, and mobile technology continue to change the world. Over his 20-year career, he’s written for Android Police, ExtremeTech, Wirecutter, NY Times, and more. He has reviewed more phones than most people will ever own. You can follow him on Bluesky, where you will see photos of his dozens of mechanical keyboards.

Google unveils Gemini 3 AI model and AI-first IDE called Antigravity Read More »

With a new company, Jeff Bezos will become a CEO again

AI, Amazon, blue origin, Jeff Bezos, project prometheus, Tech / Rejus Almole / November 17, 2025

Jeff Bezos is one of the world’s richest and most famous tech CEOs, but he hasn’t actually been a CEO of anything since 2021. That’s now changing as he takes on the role of co-CEO of a new AI company, according to a New York Times report citing three people familiar with the company.

Grandiosely named Project Prometheus (and not to be confused with the NASA project of the same name), the company will focus on using AI to pursue breakthroughs in research, engineering, manufacturing, and other fields that are dubbed part of “the physical economy”—in contrast to the software applications that are likely the first thing most people in the general public think of when they hear “AI.”

Bezos’ co-CEO will be Dr. Vik Bajaj, a chemist and physicist who previously led life sciences work at Google X, an Alphabet-backed research group that worked on speculative projects that could lead to more product categories. (For example, it developed technologies that would later underpin Google’s Waymo service.) Bajaj also worked at Verily, another Alphabet-backed research group focused on life sciences, and Foresite Labs, an incubator for new AI companies.

With a new company, Jeff Bezos will become a CEO again Read More »

Report claims that Apple has yet again put the Mac Pro “on the back burner”

Apple, Apple silicon, M2 Ultra, m3 ultra, m5 ultra, Mac Pro, Mac Studio, Tech / Kris Guyer / November 17, 2025

Do we still need a Mac Pro, though?

Regardless of what Apple does with the Mac Pro, the desktop makes less sense than ever in the Apple Silicon era. Part of the appeal of the early 2010s and the 2019 Mac Pro towers was their internal expandability, particularly with respect to storage, graphics cards, and RAM. But while the Apple Silicon Mac Pro does include six internal PCI Express slots, it supports neither RAM upgrades nor third-party GPUs from Nvidia, AMD, or Intel. Thunderbolt 5’s 120 Gbps transfer speeds are also more than fast enough to support high-speed external storage devices.

That leaves even the most powerful of power users with few practical reasons to prefer a $7,000 Mac Pro tower to a $4,000 Mac Studio. And that would be true even if both desktops used the same chip—currently, the M3 Ultra Studio comes with more and newer CPU cores, newer GPU cores, and 32GB more RAM for that price, making the comparison even more lopsided.

Mac Pro aside, the Mac should have a pretty active 2026. Every laptop other than the entry-level 14-inch MacBook Pro should get an Apple M5 upgrade, with Pro and Max chips coming for the higher-end Pros. Those chips, plus the M5 Ultra, would give Apple all the ingredients it would need to refresh the iMac, Mac mini, and Mac Studio lineups as well.

Insistent rumors also indicate that Apple will be introducing a new, lower-cost MacBook model with an iPhone-class chip inside, a device that seems made to replace the 2020 M1 MacBook Air that Apple has continued to sell via Walmart for between $600 and $650. It remains to be seen whether this new MacBook would remain a Walmart exclusive or if Apple also plans to offer the laptop through other retailers and its own store.

Report claims that Apple has yet again put the Mac Pro “on the back burner” Read More »

Google will let Android power users bypass upcoming sideloading restrictions

Apps, developers, Google, Tech / Kris Guyer / November 13, 2025

Google recently decided that the freedom afforded by Android was a bit too much and announced developer verification, a system that will require developers outside the Google Play platform to register with Google. Users and developers didn’t accept Google’s rationale and have been complaining loudly. As Google begins early access testing, it has conceded that “experienced users” should have an escape hatch.

According to Google, online scam and malware campaigns are getting more aggressive, and there’s real harm being done in spite of the platform’s sideloading scare screens. Google says it’s common for scammers to use social engineering to create a false sense of urgency, prompting users to bypass Android’s built-in protections to install malicious apps.

Google’s solution to this problem, as announced several months ago, is to force everyone making apps to verify their identities. Unverified apps won’t install on any Google-certified device once verification rolls out. Without this, the company claims malware creators can endlessly create new apps to scam people. However, the centralized nature of verification threatened to introduce numerous headaches into a process that used to be straightforward for power users.

This isn’t the first time Google has had to pull back on its plans. Each time the company releases a new tidbit about verification, it compromises a little more. Previously, it confirmed that a free verification option would be available for hobbyists and students who wanted to install apps on a small number of devices. It also conceded that installation over ADB via a connected computer would still be allowed.

Now, Google has had to acknowledge that its plans for verification are causing major backlash among developers and people who know what an APK is. So there will be an alternative, but we don’t know how it will work just yet.

How high is your risk tolerance?

Google’s latest verification update explains that the company has received a lot of feedback from users and developers who want to be able to sideload without worrying about verification status. For those with “higher risk tolerance,” Google is exploring ways to make that happen. This is a partial victory for power users, but the nature of Google’s “advanced flow” for sideloading is murky.

Google will let Android power users bypass upcoming sideloading restrictions Read More »

Microsoft releases update-fixing update for update-eligible Windows 10 PCs

Tech, Windows 10, windows 10 22h2 / Mike M. / November 12, 2025

Officially, Windows 10 died last month, a little over a decade after its initial release. But the old operating system’s enduring popularity has prompted Microsoft to promise between one and three years of Extended Security Updates (ESUs) for many Windows 10 PCs. For individuals with Windows 10 PCs, it’s relatively easy to get an additional year of updates at no cost.

Or at least, it’s supposed to be. Bugs initially identified by Windows Latest were keeping some Windows 10 PCs from successfully enrolling in the ESU program, preventing those PCs from signing up to grab the free updates. And because each Windows 10 PC has to be manually enrolled in the program, a broken enrollment process also meant broken security updates.

To fix the problems, Microsoft released an update for Windows 10 22H2 (KB5071959) this week that both acknowledges and fixes an issue “where the enrollment wizard may fail during enrollment.” It’s being offered to all Windows 10 PCs regardless of whether they’re enrolled in the ESU program “as it resolves an issue that was preventing affected customers from receiving essential security updates.”

Microsoft releases update-fixing update for update-eligible Windows 10 PCs Read More »

The Mac calculator’s original design came from letting Steve Jobs play with menus for ten minutes

1982, 1984, Andy Hertzfeld, Apple, calculator, Chris Espinosa, classic mac, computer history, Folklore.org, Mac OS, Macintosh, QuickDraw, retrocomputing, retrotech, software development, Steve Jobs, tales of tech, tales of tech history, tales of tech lore, Tech, user interface design, Vintage computing / Mike M. / November 11, 2025

Rather than continue the endless revision cycle, Espinosa took a different approach. According to Hertzfeld, Espinosa created a program that exposed every visual parameter of the calculator through pull-down menus: line thickness, button sizes, background patterns, and more. When Jobs sat down with it, he spent about ten minutes adjusting settings until he found a combination he liked.

The approach worked. When given direct control over the parameters rather than having to articulate his preferences verbally, Jobs quickly arrived at a design he was satisfied with. Hertzfeld notes that he implemented the calculator’s UI a few months later using Jobs’s parameter choices from that ten-minute session, while Donn Denman, another member of the Macintosh team, handled the mathematical functions.

That ten-minute session produced the calculator design that shipped with the Mac in 1984 and remained virtually unchanged through Mac OS 9, when Apple discontinued that OS in 2001. Apple replaced it in Mac OS X with a new design, ending the calculator’s 17-year run as the primary calculator interface for the Mac.

Why it worked

Espinosa’s Construction Set was an early example of what would later become common in software development: visual and parameterized design tools. In 1982, when most computers displayed monochrome text, the idea of letting someone fine-tune visual parameters through interactive controls without programming was fairly forward-thinking. Later, tools like HyperCard would formalize this kind of idea into a complete visual application framework.

The primitive calculator design tool also revealed something about Jobs’s management process. He knew what he wanted when he saw it, but he perhaps struggled to articulate it at times. By giving him direct manipulation ability, Espinosa did an end-run around that communication problem entirely. Later on, when he returned to Apple in the late 1990s, Jobs would famously insist on judging products by using them directly rather than through canned PowerPoint demos or lists of specifications.

The longevity of Jobs’s ten-minute design session suggests the approach worked. The calculator survived nearly two decades of Mac OS updates, outlasting many more elaborate interface elements. What started as a workaround became one of the Mac’s most simple but enduring designs.

By the way, if you want to try the original Mac OS calculator yourself, you can run various antique versions of the operating system in your browser thanks to the Infinite Mac website.

The Mac calculator’s original design came from letting Steve Jobs play with menus for ten minutes Read More »

Google says new cloud-based “Private AI Compute” is just as secure as local processing

AI, Artificial Intelligence, Google, Google Gemini, NPU, Tech / Tim Belzer / November 11, 2025

NPUs can’t do it all, though. While Gemini Nano is getting more capable, it can’t compete with models that run on massive, high-wattage servers. That might be why some AI features, like the temporarily unavailable Daily Brief, don’t do much on the Pixels. Magic Cue, which surfaces personal data based on screen context, is probably in a similar place. Google now says that Magic Cue will get “even more helpful” thanks to the Private AI Compute system.

Pixel 10 flat — Magic Cue debuted on the Pixel 10, but it doesn’t do much yet. Credit: Ryan Whitwam

Google has also released a Pixel feature drop today, but there aren’t many new features of note (unless you’ve been hankering for Wicked themes). As part of the update, Magic Cue will begin using the Private AI Compute system to generate suggestions. The more powerful model might be able to tease out more actionable details from your data. Google also notes the Recorder app will be able to summarize in more languages thanks to the secure cloud.

So what Google is saying here is that more of your data is being offloaded to the cloud so that Magic Cue can generate useful suggestions, which would be a change. Since launch, we’ve only seen Magic Cue appear a handful of times, and it’s not offering anything interesting when it does.

There are still reasons to use local AI, even if the cloud system has “the same security and privacy assurances,” as Google claims. An NPU offers superior latency because your data doesn’t have to go anywhere, and it’s more reliable, as AI features will still work without an Internet connection. Google believes this hybrid approach is the way forward for generative AI, which requires significant processing even for seemingly simple tasks. We can expect to see more AI features reaching out to Google’s secure cloud soon.

Google says new cloud-based “Private AI Compute” is just as secure as local processing Read More »

Ryanair tries forcing app downloads by eliminating paper boarding passes

airlines, Europe, Policy, Software, Tech / Tim Belzer / November 11, 2025

The policy change is also meant to get people to do more with Ryanair’s app, like order food and drinks, view real-time flight information, and receive notifications during delays.

Brady said Ryanair selected the November 12 start date because it’s during what is “traditionally a slightly quieter time for travel following the busy mid-term break period.”

Inconvenience preparing for takeoff

Eliminating paper boarding passes may create numerous inconveniences. To start, not everyone wants Ryanair’s app on their personal device. And many future customers, especially those who don’t fly with Ryanair frequently or who don’t fly much at all, may be unaware of the change, creating confusion during travel, which can already be inherently stressful.

Also, there are places where Ryanair flies that don’t accept digital boarding passes, including some airports in Albania and Morocco. In these instances, Ryanair still requires online check-ins (either via Ryanair’s website or app), and then the airline will provide paper boarding passes.

People who are less technically savvy or who don’t have a smart device or whose device has died won’t be completely out of luck. Ryanair says it will accommodate people without access to a smartphone with “a free of charge boarding pass at the airport” if they’ve checked in online “before arriving at the airport.”

“Nobody’s going to get stranded. Nobody’s going to get left behind,” O’Leary said, per The Telegraph.

Despite this, some travel experts are worried about potential chaos.

“There will be absolute havoc when that takes effect,” Irish travel commentator Eoghan Corry told RSVP Live in January.

Ryanair will be the first airline to do away with paper boarding passes. It already has a history with being on the ground floor of controversial digital-forward policies. Ryanair was the first airline to require people to check in online in advance or else pay a fee, as noted by The Independent.

“There’ll be some teething problems,” O’Leary said of moving to digital-only boarding passes.

Ryanair tries forcing app downloads by eliminating paper boarding passes Read More »

Google announces even more AI in Photos app, powered by Nano Banana

AI, AI images, Artificial Intelligence, Google, google photos, Tech / Rejus Almole / November 11, 2025

We’re running out of ways to tell you that Google is releasing more generative AI features, but that’s what’s happening in Google Photos today. The Big G is finally making good on its promise to add its market-leading Nano Banana image-editing model to the app. The model powers a couple of features, and it’s not just for Google’s Android platform. Nano Banana edits are also coming to the iOS version of the app.

Nano Banana started making waves when it appeared earlier this year as an unbranded demo. You simply feed the model an image and tell it what edits you want to see. Google said Nano Banana was destined for the Photos app back in October, but it’s only now beginning the rollout. The Photos app already had conversational editing in the “Help Me Edit” feature, but it was running an older non-fruit model that produced inferior results. Nano Banana editing will produce AI slop, yes, but it’s better slop.

Nano Banana in Help me edit

Google says the updated Help Me Edit feature has access to your private face groups, so you can use names in your instructions. For example, you could type “Remove Riley’s sunglasses,” and Nano Banana will identify Riley in the photo (assuming you have a person of that name saved) and make the edit without further instructions. You can also ask for more fantastical edits in Help Me Edit, changing the style of the image from top to bottom.

Google announces even more AI in Photos app, powered by Nano Banana Read More »

Apple TV execs dismiss introducing an ad tier, buying Warner Bros. Discovery

ads, Apple, Apple TV, streaming, Tech / Kris Guyer / November 10, 2025

Focused on original content

Another obvious way to grow Apple TV is through more subscribers. With talk of Warner Bros. Discovery considering a sale, it’s worth wondering if Apple TV may try to grow through acquisition. But the execs Screen International spoke with seemed focused on building out Apple TV’s library with originals. Cue noted that “at least in the timeframe that we’re thinking about right now, we’re not looking at licensing any content or adding anything to our service.”

“We’re building an all-original services; we’re not building on the back of pre-existing IP or library,” Jamie Erlicht, one of Apple’s heads of worldwide video, said.

More directly, when asked if Apple might buy Warner Bros., A24, or Disney, Cue pointed out that Apple hasn’t historically done “a lot of major acquisitions.”

“We do very small acquisitions in general, not related to Apple TV, so I don’t see that happening because we like what we’re doing,” Cue said.

Since its 2019 debut, some have questioned whether Apple TV is an authentic attempt to improve streaming options for customers, or if Apple TV is a “vanity project,” as Screen International put it, or if the service is merely a tool for getting people to buy other Apple products. Naturally, the interviewed executives claimed that the service is built on a commitment to distributing unique and premium shows and movies.

The interview provided more insight into how Apple TV leadership defines the latter. Zack Van Amburg, one of Apple’s heads of worldwide video, said:

A core tenet of everything Apple does is the notion that humanity needs to be at the center of it, and that’s everything from app design to hardware engineering, to everything in between. We try to think a little more deeply about that.

Our shows and our movies tend to be about the emotional experience, the stakes involved, even when we’re doing a comedy.

Apple TV execs dismiss introducing an ad tier, buying Warner Bros. Discovery Read More »

Commercial spyware “Landfall” ran rampant on Samsung phones for almost a year

android, malware, samsung, Security, smartphones, spyware, Tech / Tim Belzer / November 7, 2025

Before the April 2025 patch, Samsung phones had a vulnerability in their image processing library. This is a zero-click attack because the user doesn’t need to launch anything. When the system processes the malicious image for display, it extracts shared object library files from the ZIP to run the Landfall spyware. The payload also modifies the device’s SELinux policy to give Landfall expanded permissions and access to data.

Landfall flowchart — How Landfall exploits Samsung phones. Credit: Unit 42

The infected files appear to have been delivered to targets via messaging apps like WhatsApp. Unit 42 notes that Landfall’s code references several specific Samsung phones, including the Galaxy S22, Galaxy S23, Galaxy S24, Galaxy Z Flip 4, and Galaxy Z Fold 4. Once active, Landfall reaches out to a remote server with basic device information. The operators can then extract a wealth of data, like user and hardware IDs, installed apps, contacts, any files stored on the device, and browsing history. It can also activate the camera and microphone to spy on the user.

Removing the spyware is no easy feat, either. Because of its ability to manipulate SELinux policies, it can burrow deeply into the system software. It also includes several tools that help evade detection. Based on the VirusTotal submissions, Unit 42 believes Landfall was active in 2024 and early 2025 in Iraq, Iran, Turkey, and Morocco. The vulnerability may have been present in Samsung’s software from Android 13 through Android 15, the company suggests.

Unit 42 says that several naming schemes and server responses share similarities with industrial spyware developed by big cyber-intelligence firms like NSO Group and Variston. However, they cannot directly tie Landfall to any particular group. While this attack was highly targeted, the details are now in the open, and other threat actors could now employ similar methods to access unpatched devices. Anyone with a supported Samsung phone should make certain they are on the April 2025 patch or later.

Commercial spyware “Landfall” ran rampant on Samsung phones for almost a year Read More »

Higher prices, simpler streaming expected if HBO Max folds into Paramount+

HBO, HBO Max, mergers, paramount, paramount plus, streaming, Tech, warner bros. discovery / Tim Belzer / November 7, 2025

If a company acquires any form of HBO, one of its top challenges is expected to be streamlining operations while maintaining HBO’s premium brand. This could be especially difficult under a “more mainstream umbrella like Paramount+,” Alderman noted.

Streaming has already diluted the HBO brand somewhat. Through streaming, HBO is now associated with stuff from DC Comics and Cartoon Network, as well as reality shows, like 90 Day Fiancé and Naked and Afraid. Merging with Paramount+ or even Netflix could expand the HBO umbrella further.

That expanded umbrella could allow a company like Paramount to better compete against Netflix, something WBD executives have shied away from. HBO Max is “not everything for everyone in a household,” JB Perrette, WBD’s streaming president and CEO, said this spring.

“What people want from us in a world where they’ve got Netflix and Amazon [Prime Video] are those things that differentiate us,” Casey Bloys, chairman and CEO of HBO and Max content, told The Wall Street Journal in May.

A “stress test” for more streaming mergers

Aside from the impact on HBO Max subscribers, WBD’s merger talks have broad implications. A deal would open the door for much more consolidation in the streaming space, something that experts have been anticipating for some years and that addresses the boom of streaming services. Per Clark, discussions of a Paramount-WBD merger are “less about two studios joining forces and more about a stress test for future M&A.”

If WBD accepts a Paramount bid and that bid clears regulatory hurdles, it would signal that “premium content under fewer umbrellas is back in play,” Clark said.

A Paramount-WBD merger is likely to speed up consolidation among mid-tier players, like NBCUniversal, Lionsgate, and AMC, Alderman said, pointing to these companies’ interest in scaling their streaming businesses and in building differentiated portfolios to counter Netflix and Disney+’s expansive libraries.

If Paramount and WBD don’t merge, Clark expects to see more “piecemeal” strategies, such as rights-sharing, joint venture bundles, and streaming-as-a-service models.

Higher prices, simpler streaming expected if HBO Max folds into Paramount+ Read More »